文章基本信息

标题：CONCRETE CRACK DETECTION BASED MULTI-BLOCK CLBP FEATURES AND SVM CLASSIFIER
本地全文：下载
作者：RGUIG MUSTAFA ; EL AROUSSI MOHAMED
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2015
卷号：81
期号：1
出版社：Journal of Theoretical and Applied
摘要：Recently Automatic concrete crack detection has been converted to a real challenge for high performance of the inspection and diagnosis of concrete structures images. Generally, there are various noises such as irregularly illuminated conditions, shading and divots in the concrete images. Hence it is difficult to detect cracks automatically. In this paper, a novel and efficient approach based on Compound Local Binary Pattern (CLBP) using support vector machines is proposed for automatic concrete crack detection. The contributions of this paper include the following steps: (1) the proposed system starts by pre-processing the database images, smoothing their texture and enhancing any existing cracks, being followed by the extraction of descriptive features. Here each image is divided into several non-overlapping blocks and each block originates a feature vector. (2) The support vector machine (SVM) is successfully applied to determine the concrete crack image classification. The experimental results gave a 97.43% classification accuracy rate, which indicate that the proposed method is a promising tool for analysis of concrete structures images.
关键词：Concrete Crack Detection; Concrete Structure; Compound Local Binary Pattern; Support Vector Machine; Feature Extraction; Local Binary Pattern.